This is a new R01 application designed to evaluate clinically relevant determinants of the remarkably sustained and selective effects of monoamine releasers on cocaine self-administration in rhesus monkeys. Cocaine abuse remains a major drug abuse problem, and the development of effective pharmacotherapies continues to be a high priority for NIDA. In preliminary studies, we have found that chronic treatment with modest doses of the monoamine releasers amphetamine and PHENMETRAZINE produced robust and selective reductions in cocaine- vs. food-maintained responding for up to 28 days in rhesus monkeys. Key elements of our results from monkeys have been demonstrated by others in human laboratory studies and clinical trials. We believe these findings support the proposition that monoamine releasers warrant further study as candidate agonist medications for cocaine dependence. Toward that end, this application proposes two Specific Aims to evaluate contextual and pharmacologic determinants of monoamine releaser effects. Specific Aim 1 will examine effects of environmental and drug-history contexts on phenmetrazine-induced reductions in cocaine self-administration. Specific Aim 1a will evaluate phenmetrazine effects under conditions of alternative reinforcer availability. In clinical drug abuse treatment, pharmacotherapies are increasingly used in conjunction with contingency management techniques that introduce and control availability of alternative reinforcers. We propose to model this dual use of pharmacotherapies and alternative reinforcers, and we hypothesize that alternative reinforcer availability will enhance phenmetrazine-induced suppression of cocaine self- administration. Specific Aim 1b will evaluate phenmetrazine effects under conditions of extended access to and withdrawal from cocaine. In other drug classes (e.g. opioids), extended access promotes increased drug consumption, the development of physical dependence, and heightened drug reinforcement during withdrawal. Moreover, opioid agonist medications are most effective under conditions of withdrawal. We hypothesize that extended access to and withdrawal from cocaine will similarly enhance the ability of phenmetrazine to reduce cocaine self-administration. Specific Aim 2 will examine pharmacokinetic and pharmacodynamic determinants of monoamine releaser effects. Specific Aim 2a will correlate pharmacokinetic and behavioral effects of PHENDIMETRAZINE, a phenmetrazine prodrug and Schedule III stimulant. We hypothesize that phendimetrazine will retain phenmetrazine's ability to produce selective decreases in cocaine self- administration, but that phendimetrazine will have a slow onset of action that correlates with generation of the active phenmetrazine metabolite. Specific Aim 2b will evaluate a series of monoamine releasers that vary in selectivity to release dopamine and serotonin vs. norepinephrine. We hypothesize that releasers with reduced noradrenergic effects will display reduced abuse liability but retain an ability to produce selective reductions in cocaine self-administration.